— 147 — 



atioii was also determined in every single experiment. For this pur- 

 pose an atmometer was constructed, consisting of a Kitasatos 

 filter candle, filled with water and fitted in a glass bottle, also filled 

 with water, through the hole of a cork stopper (Fig. 1). (An 

 atmometer of similar construction is described byTranseau 1910). 

 The water evaporating from the surface of the filter candle is al- 

 ways replaced by suction from the bottle. The surface of the free 

 end of the candle was 24.3 cm 2 . 



In every experiment on transpiration it is very 

 important to know the width of the stomatal openings. 

 As is known the transpiration can either take place 

 through the cuticle or through the stomata. The width 

 of the stomata can be estimated by the method of 

 Molisch (Molisch 1912, see also Stein 1913). 

 If a drop of alcohol or xylol be placed on the lower 

 surface of the leaf the liquid will, when the stomata 

 are open, enter into the interstitial spaces of the leaf. ^ 

 This will then be transparent at the spot where the 

 liquid has entered. The penetrating power of xylol is 

 greater than that of alcohol. Therefore a leaf reacting 

 with alcohol will have stomata more widely open than 

 a leaf reacting only with xylol. In the tables given 

 below the fig. 0 indicates that neither xylol nor al- 

 cohol, fig. 1 that xylol, but not alcohol and fig. 2 that 

 both xylol and alcohol can enter into the leaves. In the 

 first case (0) the stomata are either closed or at any 

 rate very little open, in the third (2) the width is as 

 great as possible. 



In Vaccinium uliginosum and Betula the reaction 

 appears most often in spots. By infiltrating leaves 

 with water Neger found, that the behaviour of leaves 

 is different (Neger 1912); in some of them all the interstitial 

 spaces are filled with water and the appearance of the leaves is 

 homogeneous; on the other hand other leaves are spotted by infil- 

 tration, the infiltrated parts appearing as squares sharply limited 

 and surrounded of nerves, because leaves reacting in this manner 

 have many separate systems of interstitial rooms without com- 

 munication. The different groups of stomata not always being 

 open at the same time, only the systems with open stomata will be 

 infiltrated. 



Of the plants used for these experiments Vaccinium Vitis idaea 

 seems to have only one or two systems of interstitial spaces accord- 



